Resumen
This paper presents compelling evidence against the null hypothesis of a unitroot in series such IMACEC, consumption or GDP for the Chilean economy.This evidence builds on results derived by Chumacero (2000) that show thatwhen the scale variable is difference stationary, the returns on assets shouldpresent a behavior that is not supported by the data. Thus, care should be takenwhen using econometric models that impose unit roots based on traditionaltests.